Simulation of the Antarctic Ice Sheet during the Last Glacial Cycle

This repository contains a simulation of the Antarctic Ice Sheet (AIS) during the Last Glacial Cycle (LGC). This is a preliminary dataset and it should not be used without permission from the authors. All modelling details are described below.

Model information

This simulation was performed with the Yelmo ice-sheet model (Robinson et al., 2020), coupled through Yelmox and including the FastIsostasy glacial isostatic adjustment (GIA) model (Swierczek-Jereczek et al., 2024). The model versions used were:

• Yelmo v1.14.1-136-g84d1d7e1

• Yelmox v1.14.2-50-gc0c0730

• FastIsostasy v0.1-282-g073556a

The model configuration used:

• Horizontal resolution: 32 km.

• Sigma-coordinate system for the vertical discretization with 10 grid points.

• Depth-integrated viscosity approximation (DIVA, Robinson et al., 2022).

• Basal sliding law: regularized Coulomb (Joughin et al., 2019). Basal melt parameterization: quadratic non-local (Jourdain et al., 2020).

• Surface melt parameterization: insolation-temperature melt (ITM) method (Robinson et al., 2010).

• Calving: von Mises stress aproach using level set method (Morlighem et al., 2016).

• Subglacial hydrology: Bueler and van Pelt (2015).

Initialization

The AIS is initialized from present-day observed topography and ice velocity fields (Schaffer et al., 2016; Rignot et al., 2011), and is spun up under present-day boundary conditions. During the first 7.5 kyr of the spinup, the spatially explicit basal friction coefficient and the thermal-forcing correction at each basin are obtained through an optimization process so that the modeled ice thickness matches the observed present-day state, following Lipscomb et al. (2021). During the last 7.5 kyr of the spinup the optimized fields are held constant and the simulation is run towards a steady state.

Forcing

The AIS is forced through the LGC using transient temperature and sea-level boundary conditions. The temperature forcing follows a snapshot-index method:

$ T^{atm} = T_{PD}^{atm} + \alpha (t) \cdot \Delta T^{atm} $,

where $T_{PD}^{atm}$ is the present-day temperature field and $\Delta T^{atm}$ is the multi-model mean Last Glacial Maximum (LGM) minus pre-industrial atmospheric temperature anomaly from PMIP3 models. $\alpha (t)$ is a time-dependent index derived from EPICA Dome C and WAIS Divide ice-core records (Jouzel et al., 2007; Cuffey et al., 2016).

The ocean temperature anomaly is calculated as a fraction of the atmospheric anomaly:

$ \Delta T^{ocn} = f^{ocn} \cdot \Delta T^{atm} $,

such that the ocean temperature evolves analogously as

$ T^{ocn} = T_{PD}^{ocn} + \alpha (t) \cdot \Delta T^{ocn} $.

Here, $f^{ocn}$ is fixed to 0.40 , which provides close agreement with reconstructions of past Southern Ocean temperatures (Chandler and Langebroek, 2024).

Sea-level forcing is prescribed from Waelbroeck et al. (2002).

Data provided

yelmo_Antarctica_lgp_32KM.nml: The namelist configuration file containing the parameters used for this simulation.

yelmo2D.nc:  NetCDF file containing the 2D and 3D spatio-temporal output of the model. These variables are described in the table below